Drift alignment or flexure?

[swag72]

I am routinely getting 20 minute subs, with round stars and no problems. When I tried 30 min subs a while ago, the stars all got rather eggy. At the time I thought maybe that was down to flexure.

My question is whether, if it was flexure, it would be showing itself in the shorter 20 min subs as well? In which case I think I may benefit from a drift align.

I would welcome your comments on this - Would flexure be showing up with the shorter 20 min subs as well?

[Uranium235]

I get flex in as little as 6-7 min despite having a flat PHD graph, thats why ive just brought an oag, fingers crossed that will fix it.

30min frames are a bit tough with any mount, a lot can go wrong in half an hour - and any minor blips along the way will have a cumulative effect. But if you can get 20min already, you may as well stick with that becuase thats still rather a long sub length

[Tim]

Sara, you should drift align anyway.

But separate guidescopes nearly always suffer from flexure to one degree or another.

There is a method you can determine the cause of the eggy stars, detailed here:

http://deepskystacker.wikispaces.com/Measuring+differential+flexure

[swag72]

Cheers Tim, I'll look at that.

I just assumed that flexure would show up well before the 30 minute stage and so I'd be having problems at the 20 minute sub stage, but they are OK.

[George]

Sara, you should drift align anyway.

But separate guidescopes nearly always suffer from flexure to one degree or another.

There is a method you can determine the cause of the eggy stars, detailed here:

http://deepskystacker.wikispaces.com/Measuring+differential+flexure

Interesting link Tim....thanks very much.

[Gina]

Sara, you should drift align anyway.

But separate guidescopes nearly always suffer from flexure to one degree or another.

There is a method you can determine the cause of the eggy stars, detailed here:

http://deepskystacker.wikispaces.com/Measuring+differential+flexure

Thanks for the link - I'll give that a read

[swag72]

Cheers all, very useful info there.

[themos]

Flexure will affect all stars in your image equally.

If you look at that wikispaces link, you will see the Angle for each frame creeping up. That's field rotation. The centre of rotation will be where your guide star is. Knowing where your guide star is relative to your sub's centre should give you an idea of how much smearing field rotation will produce. It should be least in the part of your sub closest to the guide star and most in the part faarthest away. In numbers:

0.02 degrees rotation = 0.00035 radians (divide by 180 and multiply by pi)

So, at 2000 pixels away from the guide star you'd get 0.7 pixel smearing.

So, uniform "egginess" -> DF, non-uniform egginess -> FR

[lukebl]

Thanks for that link Tim. Very useful

...uniform "egginess" -> DF, non-uniform egginess -> FR

Brilliant. That has to be the most succunct and easily-understood explanation for eggy stars, ever! Something everyone starting out in imaging should learn as a basic rule.

[themos]

Well, it ignores non-uniform egginess which is down to optical aberrations (and is consistent from one mount setup to the next), but what the hell! Anyway, you are supposed to compare high declination shots with low declination shots to get that variable isolated.

[swag72]

Excellent explanation - At the 30min stage I have non uniform egginess - So rotation more than likely. What can you do to eliminate field rotation? I use a finder / guider which is not moveable in guiderings or anything, so quite where it's pointing I have no idea!! I though that 'in the general direction' was good enough, but clearly not.

So, ways to help combat field rotation? I have recently checked my polar alignment with the EQMOD polar align tool, so I am quite happy that this is pretty spot on. Is there anything else that can be done apart from just PA?

Edited February 14, 2012 by swag72

[Uranium235]

Thanks for that link Tim. Very useful

Brilliant. That has to be the most succunct and easily-understood explanation for eggy stars, ever! Something everyone starting out in imaging should learn as a basic rule.

Here here!

Maybe there should be a sticky (with pics) in regard to star shapes.... hmmm i believe somebody was already working on something similar.

Sara, I believe PA is the only cause of rotation. But Ive seen a similar looking effect from a cheap 0.5x reducer.

Edited February 14, 2012 by Uranium235

[themos]

Sara, I'm going to be boring and say MEASURE your Polar alignment error!

Easy way to do it: plop a camera+lens that gives you a few degrees field of view on the mount. Point the mount at the Parked position where it points roughly to Polaris. take two exposures of the North Celestial Pole region at different RA positions (make them about 90 degrees apart ) and keep those as a log of your PA for that night. Send them to me and I'll tell you how bad your PA is.

edited

Edited February 14, 2012 by themos

[Black Knight]

Slightly off topic, and apologies for hijacking your thread Sara, but can anyone tell me what degree of field rotation would be acceptable? Should I be aiming for 0 degrees rotation? In the example on the DSS wiki page linked earlier, field rotation is shown as 0.02 degrees. Would that be acceptable, or too much?

[themos]

Slightly off topic, and apologies for hijacking your thread Sara, but can anyone tell me what degree of field rotation would be acceptable? Should I be aiming for 0 degrees rotation? In the example on the DSS wiki page linked earlier, field rotation is shown as 0.02 degrees. Would that be acceptable, or too much?

I explained above, using numbers. It depends on the distance of your guide star from the centre of the image.

[Black Knight]

I explained above, using numbers. It depends on the distance of your guide star from the centre of the image.

I saw that. I understand the first line no problem. But could you explain the second line please? I'm sorry, but maths is not my strongest suit. At all!

0.02 degrees rotation = 0.00035 radians (divide by 180 and multiply by pi)

So, at 2000 pixels away from the guide star you'd get 0.7 pixel smearing.

Edited February 14, 2012 by Black Knight

[themos]

Well, guiding makes the guide star stay still and everything else rotate around it. In the DSS stack we only got the .02 degree rotation after many exposures so during one exposure the rotation must have been less than 0.01 degrees. That is 0.00017 radians (half of the value above). That rotation would move something 100 pixels away from the guide star by 0.017 pixels which is way too small to be detected. Something 1000 pixels away would be 0.17 pixels, still too small. Something 5000 pixels away, the movement would be 0.87 pixels, on the edge of detectability I might say.

So the question is how many pixels away from the guide star is your centre of the image? That's where image scale comes in. If you are shooting at 1 arcsecond per pixel that would be 5000 arcseconds away which is one degree and 23 arcminutes. Guide stars can be farther than that , depends on your setup.

I'm not sure what else needs better explaining, don't be afraid to ask.

[Black Knight]

Thanks for that Themos. I think I understand now. Thanks for taking the time to explain it, I'm very grateful!

[Tim]

Lots of useful easy to understand info in this thread, think i'll make it sticky for a while.

Thanks Themos

[ollypenrice]

Are you sure you want perfect polar alignment? There are two advantages to not having it and one disadvantage. The disadvantage is field rotation in long subs. Agreed. The advantages, though, are that all your dec corrections go the same way so you can disable one of the guide directions and reduce spurious corrections. The other is that you get natural 'dither' - ie you don't use the same pixel on the same bit of sky all night.

Flexure; I'm a sceptic. The only flexure I've ever had was with a dismally ill-designed SX OAG where the whole camera pivoted on the lock screw and needed an external strap to hold it still. Flexure can get you if one of your scopes is a reflector because the mirror can move. But if both scopes are refractors I really struggle to believe that you can't make a nice stiff system. I never see any evidence of flexure in my refractor setups using aggressively tightened ST80s in fixed rings.

I've just received an Altair Astro 60mm finder-guider and I like this becuase it has no focuser, an idea I've long recommended. You just slide the camera nosepiece up and down till you are about right and lock it. Craig Stark says you don't need good focus for guiding. Good enough for me!

I'd drift align till you can get your long subs and then forget it.

Olly

[Jessun]

I'm with Olly on this one. Imaging is all about S/N ratio so why even bother with 30 min subs Sara? The atmosphere will inevitably increase the size of the Airy disc on any star or source of photons and the gain is well I don't really know. Unless you take 20 of them? (Airy has nothing to do with air btw...)

As to perfect alignment, I believe that it can all be guided out for any reasonable sub length and if backlash is a major player with your particular mount then why not 'misalign' and 'unbalance' to keep the adjustments going in one direction using a slowish guide speed so as not to induce that pesky gear wobble?

Then dither and sigma stack and do loads of subs and that's it!

Sub length is like Watts on a car stereo IMHO...

[ollypenrice]

Sub length... well the idea is to get above the read noise with the faintest signal. I've been doing 30 min Ha subs for the first time. What I haven't done, but must do, is a back to back with 15 minute subs to see how theory and practice mesh. You still need a good stack for Sigma to work but Yves and I have a simple policy; 'as long as it takes...'

Olly

[vdb]

These formula's will help you to calculate how accurate one needs to be aligned ...

On Polar Axis Alignment

Yves

[swag72]

I have some potential issues that I am trying to sort and so will carry on in thi sthread as it is relevant and may help later down the line.

I have great 3s subs and nice round stars. When I get to 300s subs, I have awful eggy stars, but predominantly on the right hand side of the image. I know that it's not focuser alignment issues though as at 3s it's all fine.

So, in my frustration with it all, all my 300s subs are deleted!! But I did have 4 x 60s subs left. I put them through DSS as per the tutorial linked by Tim earlier in the thread and have the following DSS output screen.

How can I determine anything from this? In order to determine drift I need to divide the number of pixels by the drift figure (Dy and Dx) - How do I know the number of pixels? I was using my 314L.

Edited May 4, 2012 by swag72

[themos]

I don't see evidence of drift, the dx and dy seem pretty random. Ok, it might be something like 1/2 pixel every 4 frames but that is not going to make you lose any sleep.

You don't need to know the number of pixels in your camera. You might need to know your focal length so that you can tell how many arcseconds are seen by each pixel and figure out if the seeing conditions account for the pixel smearing you get.

Seems like a guiding issue to me, perhaps you were "chasing the seeing" that night? Or having worse than usual oscillations?